#refusal

The latest AI and developer news about #refusal, with the original source and a short summary.

Study Finds LLM Refusal Behavior Mediated by Single Direction

Researchers demonstrate that removing a single vector from the residual stream of 13 open-source models up to 72B parameters effectively disables refusal mechanisms.

·1·2026.11.08 10:00

Claude refusal patterns across effort levels

Raising effort on Opus 4.7 barely changed refusal patterns, but made answers longer.

·1·2026.04.28 05:00

The Anatomy of the Refusal Circuit

A refusal circuit with a sparse gate-amplifier structure was repeatedly confirmed across 12 open-weight models.

·2026.04.15 05:00